Research On Application Of Wavelet Theory In Electric Power Quality Analysis And Processing

Electric power quality disturbance data saving and transmitting through supervision network are two important problems concerned with kinds of electric power parameters detection and analysis online systems. Wavelet transform (WT) has property of multi-scale analysis, localization both in time and frequency domain and sensitivity to the abrupt change of signals. These merits make it a new way for non-stationary signal processing. Consequently, this paper is such an attempt to take the advantage of these merits of WT in order to open a new way for solving the two problems of electric power quality disturbance data saving and transmitting through network.Background of this paper is the development of Power Parameters Detection and Analysis Online System. Aiming at the problem of the bulk of disturbance or fault data system needs to save, which makes the system of bad performance and increases the cost of saving, a data compression way based on transform coding method of wavelet transform with little information lost is proposed. At the same time its use in data filtering is also considered. Main job done consists of three parts as following:First of all, filter banks theory of digital signal analysis is studied. Sub-band coding based on two-channel quadrature-mirror filter bank is the common way for loss less data compression. Emphasis here is the application of orthogonal filter banks and biorthogonal filter banks.Then basic wavelet theory is introduced. Discrete wavelet transform has tight relationship with filter banks in born. Multi-resolution analysis showed QMF banks are equivalent to quadrature wavelet transform. Fast wavelet transform could be achieved by cascading QMF banks operation. A new result of bonding multi-channel digital signal analysis, filter banks with wavelet transform is the lifting scheme.Finally, synthesizing multi-channel digital signal processing, filter banks and wavelet analysis theory, an electric power quality disturbance data compression realization has been proposed based on quadrature discrete wavelet transform. A practical implementation has been suggested together. A demo programmed with Delphi has also been given to do compression andde-compression with four sets of simulation data. Simulation results show that with acceptable information lost within a limitation requirement remarkable compression ration could be obtained. This data compression way could benefit electric power quality disturbance data saving and transmission, improve supervision system's performance and lower down the system cost.